If you manage customer accounts, you know “customer health” isn’t just a buzzword—it’s how you spot churn before it happens (or spot a happy customer before they become a case study). But tracking it by hand? That’s a fast track to missed signals and late nights in spreadsheets. This guide is for folks who want to move past guessing and set up automated customer health scoring in Planhat, step by step, without wasting time on features you’ll never use.
What Is Customer Health Scoring—And Why Automate It?
Let’s clear something up first: customer health scoring is just a way to measure how your customers are doing using data points (not gut feelings). When you automate it, you ditch manual tracking and get early warnings—so you can react before things go sideways.
But here’s the catch: a fancy score isn’t magic. If you feed in junk data, you get junk alerts. So, this guide walks you through what actually works, what’s fluff, and how to get Planhat doing the heavy lifting for you.
Step 1: Decide What “Healthy” Means for Your Business
Before you even log in to Planhat, figure out what “healthy” looks like for your customers. This isn’t copy-paste from another SaaS company—your metrics should match your customer journey.
What to consider: - Product usage: Are they logging in? Using key features? - Support tickets: Lots of tickets could mean trouble, but none could mean they’re not engaged. - NPS/CSAT scores: Useful, but don’t trust them alone. - Contract data: Renewals, upgrades, downgrades. - Billing status: Churn risk if payments are late or missing.
What to skip: Vanity metrics. Don’t add “website visits” or “number of emails sent” unless they truly matter for your business.
Pro tip: Talk to your CS team and ask, “What’s the first thing you check when you want to know if a customer is struggling?” Start with those signals.
Step 2: Gather the Right Data in Planhat
Now, get your data into Planhat. This step is boring, but critical—automation only works if your data’s clean and consistent.
How to do it: - Connect integrations: Plug in your product, CRM, and support systems. Planhat supports a lot, but focus on the ones you actually use daily. - Map fields: Double-check the data is going to the right spots. If “Product Usage” is really “Last Login,” make that clear. - Backfill missing data: If you’re missing months of info, fill in what you can. Otherwise, your first scores will be garbage.
Pitfall to avoid: Don’t hook up every integration just because you can. More data isn’t always better—irrelevant data just clutters things up.
Step 3: Build a Simple Health Score Model
Here’s where most guides go off the rails, suggesting you build a complex model with 15 inputs. Don’t. Start with 3-5 metrics that matter.
To build your initial model in Planhat: 1. Go to the “Health” section in Planhat. 2. Click to create a new Health Profile. 3. Name it clearly (e.g., “Core Health Score”). 4. Add metrics as rules or formula components.
For each metric: - Define thresholds (e.g., “Logged in at least once in the past week”). - Assign weights. If product usage is king, give it more weight than NPS. - Decide if you want color coding (like Red/Yellow/Green)—it’s helpful, but don’t obsess over it.
Keep it simple: If you can’t explain your health score to a new hire in two minutes, it’s too complicated.
What works: Weighted, clear rules (e.g., “If usage drops by 50%, that’s a problem.”) What doesn’t: Overly vague formulas or “black box” models you can’t explain.
Step 4: Automate Health Score Updates
By default, Planhat recalculates health scores whenever the underlying data changes. But you should double-check your automation settings.
Make sure: - Data syncs are happening regularly (ideally daily). - Any custom calculations update in real-time or on a schedule that makes sense. - Health scores trigger alerts or tasks for your CS team (no point in automating if nobody acts on it).
How to set this up: - In Planhat, go to your Health Profile settings. - Set triggers for events (e.g., “If health drops to Red, create a task.”). - Test it—change a customer’s data and confirm the score updates automatically.
Pro tip: Don’t set up a flood of alerts. Focus on the triggers that actually matter, or your team will start ignoring them.
Step 5: Review, Test, and Tweak (Relentlessly)
Even the best health model is just a guess until you see it in action. Watch what happens for a couple of weeks.
Review: - Are the right customers showing as “at risk” or “healthy”? - Are there false positives/negatives? - Is your team actually using the scores, or ignoring them?
Tweak: - Adjust weights and thresholds based on real outcomes. - Drop metrics that aren’t predictive. - Add new ones only when you have data to back them up.
What to ignore: Fancy machine learning add-ons or “AI-powered” suggestions—at least until your basic model is working. Most teams never need them.
Step 6: Share Health Scores with the Right People
There’s no point having health scores if nobody sees them. Make them visible in the places people work.
How to do it: - Add health score columns to customer lists in Planhat. - Set up dashboards for execs and CSMs. - Integrate with Slack, email, or CRM if that’s where your team lives.
Don’t: Put health scores in customer-facing portals unless you’re 100% confident. Customers don’t want to see themselves marked as “unhealthy.”
Step 7: Keep It Simple and Iterate
The best health scores are living things. They should evolve as your customers do. Once a quarter, revisit your model:
- Are the metrics still relevant?
- Has the product changed?
- Are you missing obvious warning signs?
Pro tip: Make small changes, not sweeping overhauls. If you’re constantly tinkering, something’s off with your foundation.
That’s it—no magic, no black boxes, just a step-by-step way to set up automated customer health scoring in Planhat that actually helps your team. Keep it simple. Don’t stress about perfection. Start with a basic model, automate what matters, and adjust as you go. The goal isn’t a perfect score—it’s spotting problems (and wins) before everyone else.